"isso,isso,isso,isso"

if you want something please send me a DM on twitter, or support me on my patreon(there is exclusive models there)
https://www.patreon.com/wenderinf

A model based on El Chavo del Ocho.

Due to the new site law, all other detailed examples will be shared on my patreon.
consider supporting me!

Please give star reviews with your picture creations if you like it

(he has 18+)

Use any checkpoint based on SD1.5

he might look like someone from real life, but she is not what you think. She is a totally fictional character.

I generally do not support the use of this model to generate any adult content. When using this model, users should abide by local laws and regulations and shall not violate any other's right to reputation, privacy or portrait.

"The protagonist and THE HERO"

if you want something please send me a DM on twitter, or support me on my patreon(there is exclusive models there)
https://www.patreon.com/wenderinf

A model based on Kasuga Ichiban,from Japan.

Due to the new site law, all other detailed examples will be shared on my patreon.
consider supporting me!

Please give star reviews with your picture creations if you like it

(he has 18+)

Use any checkpoint based on SD1.5

he might look like someone from real life, but she is not what you think. She is a totally fictional character.

I generally do not support the use of this model to generate any adult content. When using this model, users should abide by local laws and regulations and shall not violate any other's right to reputation, privacy or portrait.

"Need Money?"

if you want something please send me a DM on twitter, or support me on my patreon(there are exclusive models there)
https://www.patreon.com/wenderinf

A model based on Akiyama,from Japan.

Due to the new site law, all other detailed examples will be shared on my patreon.
consider supporting me!

Please give star reviews with your picture creations if you like it

(he has 18+)

Use any checkpoint based on SD1.5

he might look like someone from real life, but she is not what you think. She is a totally fictional character.

I generally do not support the use of this model to generate any adult content. When using this model, users should abide by local laws and regulations and shall not violate any other's right to reputation, privacy or portrait.

This came out of an attempt to do a venom-style symbiote, but then drifted into slime and liquid latex, so... Slymebiote.

Slymebiote clothing has a tight-fitting latex-like appearance, add words like 'dripping' and increase the weight to get wierd effects.

Trained a mix of about 40 images for 30k. (Not counting the failed attempts and restarts.)

Works better in semi-realistic models.

封面圖由anything4.5.ckpt產出

This is a stylized Hypernetwork for Cornflower v7/v8/v9/X/X Feat. Offset Noise. Of course, other models can theoretically use it, but I haven't tested them.

If you like my work you might also like my Patreon.

https://www.patreon.com/AIGeneratedEroticGirls

It's better for the "Chillout-mix" or "Kanpiromix " checkpoint, besides that I haven't tested it yet

I recommend using weights (0.6 - 0.8) for better results.

It is also recommended to mix it with "lora:Japanese-doll-likeness."

Resolution 512x512

※Please use the product in compliance with the laws of each country.

If you like my work you might also like my Patreon.

https://www.patreon.com/AIGeneratedEroticGirls

It's better for the "Chillout-mix" or "Kanpiromix " checkpoint, besides that I haven't tested it yet

I recommend using weights (0.6 - 0.8) for better results.

It is also recommended to mix it with "lora:Japanese-doll-likeness."

Resolution 512x512

※Please use the product in compliance with the laws of each country.

"The best BFF ever"

if you want something please send me a DM on twitter, or support me on my patreon(there is exclusive models there)
https://www.patreon.com/wenderinf

A model based on Goro Majima,from Japan.

Due to the new site law, all other detailed examples will be shared on my patreon.
consider supporting me!

Please give star reviews with your picture creations if you like it

(she has 18+)

Use any checkpoint based on SD1.5

She might look like someone from real life, but she is not what you think. She is a totally fictional character.

I generally do not support the use of this model to generate any adult content. When using this model, users should abide by local laws and regulations and shall not violate any other's right to reputation, privacy or portrait.

"The best dad ever"

if you want something please send me a DM on twitter, or support me on my patreon(there is exclusive models there)
https://www.patreon.com/wenderinf

A model based on Kazuma Kiryu,from Kamurocho,Japan.

Due to the new site law, all other detailed examples will be shared on my patreon.
consider supporting me!

Please give star reviews with your picture creations if you like it

(she has 18+)

Use any checkpoint based on SD1.5

She might look like someone from real life, but she is not what you think. She is a totally fictional character.

I generally do not support the use of this model to generate any adult content. When using this model, users should abide by local laws and regulations and shall not violate any other's right to reputation, privacy or portrait.

"A irmã malvada"

if you want something please send me a DM on twitter, or support me on my patreon (Gor will be a exclusive model in my patreon).
https://www.patreon.com/wenderinf

A model based on Samira,Mutant from Brazil.

CAUTION: THIS MODEL IS KINDA BAD, WILL NEED A 2.0 VERSION, USE IT WITH CARE

Due to the new site law, all other detailed examples will be shared on my patreon.
consider supporting me!

Please give star reviews with your picture creations if you like it

(she has 18+)

Use any checkpoint based on SD1.5

She might look like someone from real life, but she is not what you think. She is a totally fictional character.

I generally do not support the use of this model to generate any adult content. When using this model, users should abide by local laws and regulations and shall not violate any other's right to reputation, privacy or portrait.

Makes breasts bigger.

(IMPORTANT: In my opinion V1.0 is better than v2.x. I'm providing the 2.x file only so you can try it out if you like. Combining both versions in the same prompt will get you even bigger breasts, but 2.x causes a higher rate of trash images.)

If applied to a picture of a woman, will tend to increase the size of her breasts. Also skews images towards realism.

This will work somewhat without any specific keywords, for example:

vs

However, it responds very strongly to any mention of breasts, especially size modifiers such as "large", "huge", or "enormous":

vs

The above examples are generated with the Deliberate model based on SD1.5.

Strongest effect seems to be with the exact phrase "enormous soft natural breasts":

vs

Above example generated with the UberRealisticPornMerge model.

Other key phrases that get good results:

• Suggestions of "stretching" clothes and "bursting" or "bulging" out of clothes.

• "Detailed realistic skin texture" or variants on that phrase.

• "Soft" and "Natural"

Problems that may be fixed in future versions:

• Way too eager to get girls naked. Will straight up ignore "NSFW" and "naked" negative prompts a lot of the time, or make clothes partial or transparent. Use increased strength on negative prompts to avoid this.

• Is trained to understand prompts related to freckles, veins, areolas and nipple size, but comes on very strong at normal strength. Dial it down to 0.5 or less to get better results.

• Duplicating synonyms of "breasts" in prompts (e.g. "tits", "boobs") will tend to get additional breasts on the same woman instead of larger ones. If you're into that, go for it.

• Does not do well at complex poses. Though sometimes will actually fix weirdness generated by the base model. For example, seems to improve Deliberate's attempts at laying down poses.

Please review if you like this! Comment if you find something that doesn't work, or if you've got a use case that needs improving for the next version. Show me your images!

Additional interesting notes: this hypernetwork has unexpected interesting effects on images not of women. Landscapes just seem to look nicer? Scenes seem better composed with more interesting backgrounds? It especially seems to like water and trees, but ground textures seem to be affected positively too.

Model based on internet model Imogen Dyer

"A nova solteira do Pedaço?"

if you want something please send me a DM on twitter, or support me on my patreon (M4rvil4 will be a exclusive model in my patreon).
https://www.patreon.com/wenderinf

A model based on Key from Brazil,Volleyball Player.

Due to the new site law, all other detailed examples will be shared on my patreon.
consider supporting me!

Please give star reviews with your picture creations if you like it

(she has 18+)

Use any checkpoint based on SD1.5

She might look like someone from real life, but she is not what you think. She is a totally fictional character.

I generally do not support the use of this model to generate any adult content. When using this model, users should abide by local laws and regulations and shall not violate any other's right to reputation, privacy or portrait.

"o cara que jogou um bebê no lixo e vendeu hot dog na mesma novela"

if you want something please send me a DM on twitter, or support me on my patreon (There is a lot of exclusive models in my patreon).
https://www.patreon.com/wenderinf

A model based on M4t3us Solan0 from Brazil,Actor.

Due to the new site law, all other detailed examples will be shared on my patreon.
consider supporting me!

Please give star reviews with your picture creations if you like it

(he has 18+)

Use any checkpoint based on SD1.5

he might look like someone from real life, but he is not what you think. he is a totally fictional character.

I generally do not support the use of this model to generate any adult content. When using this model, users should abide by local laws and regulations and shall not violate any other's right to reputation, privacy or portrait.

"Amizade com Ney terminada, agora sou melhor amiga do Besouro"

if you want something please send me a DM on twitter, or support me on my patreon (There is a lot of exclusive models in my patreon).
https://www.patreon.com/wenderinf

A model based on Brun4 M4rqu3z1n3 from Brazil,Actress.

Due to the new site law, all other detailed examples will be shared on my patreon.
consider supporting me!

Please give star reviews with your picture creations if you like it

(she has 18+)

Use any checkpoint based on SD1.5

She might look like someone from real life, but she is not what you think. She is a totally fictional character.

I generally do not support the use of this model to generate any adult content. When using this model, users should abide by local laws and regulations and shall not violate any other's right to reputation, privacy or portrait.

"Você não vai flopar essa relação"

if you want something please send me a DM on twitter, or support me on my patreon (There is a lot of exclusive models in my patreon).
https://www.patreon.com/wenderinf

A model based on J4d3 P1c0n from Brazil,Actress.

Due to the new site law, all other detailed examples will be shared on my patreon.
consider supporting me!

Please give star reviews with your picture creations if you like it

(she has 18+)

Use any checkpoint based on SD1.5

She might look like someone from real life, but she is not what you think. She is a totally fictional character.

I generally do not support the use of this model to generate any adult content. When using this model, users should abide by local laws and regulations and shall not violate any other's right to reputation, privacy or portrait.

"Filha do Fabinho"

if you want something please send me a DM on twitter, or support me on my patreon (There is a lot of exclusive models in my patreon).
https://www.patreon.com/wenderinf

A model based on Cl30 P1r3s from Brazil,Actress.

Due to the new site law, all other detailed examples will be shared on my patreon.
consider supporting me!

Please give star reviews with your picture creations if you like it

(she has 18+)

Use any checkpoint based on SD1.5

She might look like someone from real life, but she is not what you think. She is a totally fictional character.

I generally do not support the use of this model to generate any adult content. When using this model, users should abide by local laws and regulations and shall not violate any other's right to reputation, privacy or portrait.

"NÃO FALA NADA MUDA"

if you want something please send me a DM on twitter, or support me on my patreon (Al4nis Marruá will be a exclusive model in my patreon).
https://www.patreon.com/wenderinf

A model based on B3LL4 C4MP0S from Brazil,Actress.

Due to the new site law, all other detailed examples will be shared on my patreon.
consider supporting me!

Please give star reviews with your picture creations if you like it

(she has 18+)

Use any checkpoint based on SD1.5

She might look like someone from real life, but she is not what you think. She is a totally fictional character.
I generally do not support the use of this model to generate any adult content. When using this model, users should abide by local laws and regulations and shall not violate any other's right to reputation, privacy or portrait.

"CANETA AZUL, AZUL CANETA, CANETA AZUL TA GERADO NA MINHA TELA"

My april Fools model, hope you guys like it! FELIZ 1 DE ABRIL

Consider donating to my Patreon(there is exclusive content and models there!):
https://www.patreon.com/wenderinf

A model based on Man0el G0meZ THE BLUEPEN.

Due to the new site law, all other detailed examples will be shared on my patreon.

consider supporting me! AND DON'T FORGET TO REVIEW!

Use any checkpoint based on SD1.5

he might look like someone from real life, but she is not what you think. She is a totally fictional character.

Transform the images into Girls Frontline chibi style.

Trigger Words: smile, transparent_background, chibi, <hypernet:GFL-2000:1>

"TIROU A PERUCA DO SILVIO!"

if you want something please send me a DM on twitter, or support me on my patreon (L4r1ss4 M4noela will be a exclusive model in my patreon). https://www.patreon.com/wenderinf

A model based on +A from Brazil,Actress. Due to the new site law, all other detailed examples will be shared on my patreon.
consider supporting me!

Please give star reviews with your picture creations if you like it

(she has 18+)

Use any checkpoint based on SD1.5

She might look like someone from real life, but she is not what you think. She is a totally fictional character.

"PREPARA!"
if you want something please send me a DM on twitter, or support me on my patreon (Pabbl0 V1tt4r will be a exclusive model in my patreon).
https://www.patreon.com/wenderinf

A model based on Ann1t4 from Brazil,Singer.

Due to the new site law, all other detailed examples will be shared on other social medias

Please give star reviews with your picture creations if you like it

(she has 18+)

Use any checkpoint based on SD1.5

She might look like someone from real life, but she is not what you think. She is a totally fictional character.

curaの約400枚の画像で学習したhypernetworkです。

This is a hypernetwork trained on about 400 images of cura.

curaの代表作:まいてつ、ものべの

cura's best-known works: maitetsu(https://maitetsucs.com/), monobeno

danbooru:https://danbooru.donmai.us/posts?tags=cura&z=5

base model:Anything-v4.5

curaの画風の再現には成功しているように見えますが、特定のキャラクターの再現性は低いです(キャラクター再現は目的ではありません。

It seems to be successful in reproducing cura's painting style, but the reproduction of specific characters is poor (character reproduction is not the goal.

Lowdef

Lowdef is a model trained on a stylized lowpoly dataset that captures a unique low-definition style (base model SD 1.5). It's not meant to be used stand-alone but with other checkpoints.

Auto1111 Quick Start

Instructions for use with Stable Diffusion Web UI.

1. Download the model. Rename the file to lowdef.pt or alternatively, adjust your prompts to match the file name.

2. Place the downloaded file inside stable-diffusion-webui/models/hypernetworks directory. If the hypernetworks directory doesn't exist create it.

3. Add <hypernet:lowdef:0.25> to your prompt and adjust the blend to your liking.

Example

Your Prompt <hypernet:lowdef:0.25>

Best Practices

⚠️ The model is quite aggressive and more unpredictable at higher blend values.

• Use a blend between 0.1 and 0.3.

• Generate multiple images at once, minimum 4.

• Use Lowdef with other artistic checkpoints.

"BOTA,BOTA, NA P1P0K1NH4"
if you want something please send me a DM on twitter, or support me on my patreon (there is exclusive models in my patreon).
https://www.patreon.com/wenderinf

A model based on P1P0K1NHA from Brazil,Singer and OF model.

I recomend that you use this one with more care, can generate some cursed results because her tooth braces.

Due to the new site law, all other detailed examples will be shared on other social medias

Please give star reviews with your picture creations if you like it

(she has 18+)

Use any checkpoint based on SD1.5

She might look like someone from real life, but she is not what you think. She is a totally fictional character.

if you want something please send me a DM on twitter, or support me on my patreon (there is exclusive models in my patreon).
https://www.patreon.com/wenderinf

A model based on P4olla Oliv3ir4 from Brazil,Actress and Fitness Model.

I recomend that you put black and white on the negative, because this model can generate some B&W pictures.

Due to the new site law, all other detailed examples will be shared on other social medias

Please give star reviews with your picture creations if you like it

(she has 18+)

Use any checkpoint based on SD1.5

She might look like someone from real life, but she is not what you think. She is a totally fictional character.

This is a helper plugin, hyper-network. It's mostly made to be used alongside my model Gloomsday. Check my profile to find the model link. It greatly enhances the art. I highly recommend you use it for most of the art in Gloomsday.

No images of it, just download it and place it in your hypernetwork folder for usage. Place the tag at the start of your prompt for the effect to work.

if you want something please send me a DM on twitter, or support me on my patreon (her best friend will be exclusive for my patreon).
https://www.patreon.com/wenderinf

A model based on Brun4 Gr1ph40 from Brazil.

Due to the new site law, all other detailed examples will be shared on other social medias

Please give star reviews with your picture creations if you like it

(she has 18+)

Use any checkpoint based on SD1.5

She might look like someone from real life, but she is not what you think. She is a totally fictional character.

if you want something please send me a DM on twitter, or support me on my patreon (i will launch exclusive content there).
https://www.patreon.com/wenderinf

A model based on Iz4 from Brazil.

Due to the new site law, all other detailed examples will be shared on other social medias

Please give star reviews with your picture creations if you like it

(she has 18+)

Use any checkpoint based on SD1.5

She might look like someone from real life, but she is not what you think. She is a totally fictional character.

if you want something please send me a DM on twitter, or support me on my patreon (her best friend will be exclusive for my patreon).
https://www.patreon.com/wenderinf

This was a request, if you want something please send me a DM on my twitter @wenderinf.

A model based on Shir4K1n

Due to the new site law, all other detailed examples will be shared on other social medias

Please give star reviews with your picture creations if you like it

(she has 18+)

Use any checkpoint based on SD1.5

She might look like someone from real life, but she is not what you think. She is a totally fictional character.

if you want something please send me a DM on twitter, or support me on my patreon (i will launch exclusive content there).
https://www.patreon.com/wenderinf

A model based on Yum3r1

Due to the new site law, all other detailed examples will be shared on other social medias

Please give star reviews with your picture creation if you like it

(she has 18+ years despite her looks)

Use any checkpoint based on SD1.5

She might look like someone from real life, but she is not what you think. She is a totally fictional character.

Straight out of the 90s, when comic books were solely marketed to teen boys and realistic anatomy was a foreign concept, comes Witchblade. Woman finds a glove that gives her superpowers along with wildly impracticable 'armor'. Somehow it also got a live-action (fully clothed) TV show on TNT.

Witchblade is a thing, so "woman wearing witchblade" works best. Weight at 1.0. Lower it if you want more coverage. Can be asymmetrical at times due to source material. It's a feature, not a bug.

Works good on non-anime models. Anything-v3 usually makes something closer to Kiki's Delivery Service.

This is my first time training a model, but i hope you enjoy:

Dont forget to use the tags: full_body, virtual_youtuber, standing

Female tongues and mouths

Model intended to generate more realistic tongues and mouths.

Can produce mouths and tongues of any shapes or sizes, photorealistic if desired. Given the appropriate prompt, long tongues from likely human to anything.

Trained on SD 1.5, tested with models Deliberate and Dreamshaper

Provided samples include base SD 1.5 model, Deliberate and Dreamshaper. These are RAW samples that have not been improved with GFPGAN or CodeFormer. Provided samples of realistic output, fantasy, humongous, etc.

Triggers: tongue, tongue out, mouth open

More samples on https://www.deviantart.com/znort2005

テラスMCの画像で学習したハイパーネットワークです。

Hypernetwork trained on terasu MC images.

pixiv:https://www.pixiv.net/users/11701235 (NTR注意)

danbooru:https://danbooru.donmai.us/posts?tags=terasu_mc+&z=5

base moel:Anything-v4.5

画風の表現には成功しているように見えますが、学習に使用した画像に漫画的なコマ分けや複数人の複雑な構図が多いためか、プロンプトで指定した構図や違和感のない人間の生成がうまくできないことが多いです。

Although it seems to be successful in expressing the painting style, it often fails to generate the composition and uncomfortable people specified in the prompts, perhaps because the images used for training have many cartoonish frame divisions and complex compositions of multiple people.

THIS IS A PROOF OF CONCEPT

This hypernetwork lets you generate landscapes that look like Minecraft with shaders. That's the whole point kinda.

The training image pool was pretty scarce for the first version, and the images end up looking very repetitive. Still, pretty cool for an experiment.

This is a hypernetwork model of a painter

For reasons of copyright and reputational rights, I will not point out the artist and my protection here

And the version provided is not highly trained

(But actually, it works and plays a certain style of rendering work, you can run it alone, or in combination with other lora or hyp files)

PS: This model is rendered by anything 2.0, and it is recommended to generate 512x768

这是一个画师的超网络模型

出于版权以及名誉权的考虑，对画师以及和我的保护，我在此并不会指出画师

且提供的版本并非高度训练后的（我留了训练更高宽度和深度的，但我觉得这个模型已经足够作为画风渲染了）

（但实际上， 它可以运行某种风格的渲染工作， 你可以单独运行它， 或与其他 lora 或 hyp 文件组合使用）

PS：该模型由 anything 2.0进行渲染，推荐生成512x768

Asari from the Mass Effect series. Embrace eternity or something like that. (That whole Consort plotline kinda went nowhere, didn't it?)

Trained on SD1.5 and can get surprisingly good results there. It doesn't get along too well with Anything V3. Tends to bleed Mass Effect style into your settings.

if you want something please send me a DM on twitter, or support me on my patreon (i will launch exclusive content there).
https://www.patreon.com/wenderinf

A model based on An4nd4

Due to the new site law, all other detailed examples will be shared on other social medias

Please give star reviews with your picture creation if you like it

(she has 18+ years despite her looks)

Use any checkpoint based on SD1.5

She might look like someone from real life, but she is not what you think. She is a totally fictional character.

"OOOOOOIIII AMOOOOOOR"

if you want something please send me a DM on twitter, or support me on my patreon (i will launch exclusive content there).
https://www.patreon.com/wenderinf

A model based on Jul1an4 B0nd3

Due to the new site law, all other detailed examples will be shared on other social medias

Please give star reviews with your picture creation if you like it

(she has 18+)

Use any checkpoint based on SD1.5

She might look like someone from real life, but she is not what you think. She is a totally fictional character.

if you want something please send me a DM on twitter, or support me on my patreon (i will launch exclusive content there).
https://www.patreon.com/wenderinf

A model based on Sayur1, yes, she is a transgender girl

Due to the new site law, all other detailed examples will be shared on other social medias

Please give star reviews with your picture creation if you like it

(she has 18+ years despite her looks)

Use any checkpoint based on SD1.5

She might look like someone from real life, but she is not what you think. She is a totally fictional character.

if you want something please send me a DM on twitter, or support me on my patreon (i will launch exclusive content there).
https://www.patreon.com/wenderinf

A model based on Alessandr4 N3gr1ni

Due to the new site law, all other detailed examples will be shared on other social medias

Please give star reviews with your picture creation if you like it

(she has 40+ despite her looks)

Use any checkpoint based on SD1.5

She might look like someone from real life, but she is not what you think. She is a totally fictional character.

if you want something please send me a DM on twitter, or support me on my patreon (i will launch exclusive content there).
https://www.patreon.com/wenderinf

A model based on N3mot0

Due to the new site law, all other detailed examples will be shared on other social medias

Please give star reviews with your picture creation if you like it

(she has 18+ )

Use any checkpoint based on SD1.5

She might look like someone from real life, but she is not what you think. She is a totally fictional character.

Hypernetwork of adult Rosemary Winters from RE:Village. Used edgeofRealism to train and generate.

500枚のえびふりゃの画像で学習したハイパーネットワークです。

This is a hypernetwork trained on 500 Ebifurya images.

pixiv:https://www.pixiv.net/users/11424595

danbooru:https://danbooru.donmai.us/posts?tags=ebifurya&z=5

basemodel:Anything-v4.5

"ebifurya"などのタグを入れると効果が上がるかもしれません。他にも学習に艦これキャラの画像を多く含んでいるため、"kantai collection"や"kancolle"などのタグも影響を受けている可能性があります。

Tags such as "ebifurya" may be added to improve the effect. Other tags such as "kantai collection" and "kancolle" may also be affected, since the study contains many images of kancolle characters.

刚出炉的风格模型，这个模型风格可以制作自己喜欢的头像，禁止商用！！！

可以结合一些优质的底模和LORA，比如

dalcefoPainting_v4

anypastelAnythingV45_anypastelAnythingV45

Blue Mary (The King of Fighters) LoRA

Ellen Ripley is trained on the Delegate v2 model. Screenshots from movies were used for training.

トリガーワード無し
10000~14000がおすすめです

Nendoroid Model trainined over 250 images.

The phased used to train the model was: "Nendroid <Anime> <Character> Figure", i can be added as single tag like the examples.

Good results for me are at 0.7

Support me: https://boosty.to/supern13378

HEY LOOK, THE KING OF CHIBI!

learned Maplestory2 art. please remember dead game

recommend weights 0.8~, (chibi:1.2~1.4), and full body

useful for creating not chibi character,. Higher weights make fingers unstable

shiny gold bodypaint

use a weight of 0.7-0.9 and gold bodypaint to help it along

trained against deliberate_v11

v2.0 is slightly more general purpose as it supports clothes better

v1.0 is best for nudes

This hypernetwork learned 990 images of matsunaga kouyou.

pixiv:https://www.pixiv.net/users/440400

danbooru:https://danbooru.donmai.us/posts?tags=matsunaga_kouyou+&z=5

base model:Anything-v4.5(https://huggingface.co/andite/anything-v4.0/blob/main/anything-v4.5.safetensors)

No target tag is required; just load the hypernetwork and generate the image and the style will be applied.

This is my first learned work. After generating several images, it seems that it is still not good at drawing hands.

this model trained on personalami arts

Best results with Anything Furry model and 0.75 hypernetwork strength

You can try different samplers (Euler a, DPM++ 2* Karass and similar)

You can try to combine with Corneo's Embeddings for better results.

⚠!!! some of the Samples was upscaled with SwinIR_4x upscaler. maybe not the best idea. Check results without upscaler.

Next version should be better

borshig_personalami 💓

cetusmix is best to output

If u use upscaler, use esrgan4x or esrgan4x anime6b

will be release with a regularized model

Trained SD v1.5

maybe difficult to harmonize with background

https://twitter.com/Akina_sd

Personal Hypernetwork made of Liz Vicious, an alt model active around 2005-2007. No real reason other than I always liked her look.

There was no shortage of material to train off of, but I limited to about 174 images.

The sample images were just simple prompts I used to give an example of what I'm going for.

I might retrain it later due to not really knowing how to make these that well.

Hypernetwork of the sexy villainess from the Final Fantasy 7 Remake.

Trained off of 100 images of both screenshots from the game and some "fan made" animations and images.

Haven't done a lot of tinkering with it. But without prompts it tends to give her more grey hair than blonde. Probably due to RedMoa's animations.

It does also work in both anime style and realistic checkpoints, though realistic ones will often look more 3D rendered.

I'm trying to recreate some of the art from board game "Unstable Unicorns", starting with the babies, but it seems kinda hard to get some variations that fit with the game artstyle.

It is my first try, so any feedback is helpful.

It works better with square images, it also works fine with many styles (last images) besides the "base" style.

A bit overcooked maybe, it works better around: 0.7 ~ 0.9

You totally need: "unicorn, cartoon, baby unicorn" in your prompt to keep the base style, you can remove it to make something with another style.

Optional: "horn, wings, hickey"

Mana's youtube
Trained on Anything 4.5

promts I usualy use
+
1girl, ahoge, bangs, blue_hair, brown_eyes, choker, hair_ornament, hoodie, hood down, looking_at_viewer, short_hair, collar, virtual_youtuber, white_background
-
nsfw, lowres, bad_anatomy, bad_hands, text, error, missing_fingers, extra_digit, fewer_digits, cropped, worst_quality, low_quality, normal_quality, jpeg_artifacts, signature, watermark, username, blurry, artist_name

Does she really need an introduction? It's Mai Shuranai. She wields a folding fan and fan service. Seems to work well with all models, but leans towards anime/drawn. Can be NSFW due to wardrobe malfunctions.

'maishu' is the trigger.

Personal hypernetwork made to look like the pornstar Veruca James.

Trained with 245 images using 1.5 base.

Tested a few checkpoints, best results with URPM for realistic and Galena Blend or Sardonyx Blend for illustrated.

oov1wq legacy hypernetwork

This is not my work and I don't plan to support this, just trying to preserve this Hypernetwork for the community.

What does it do?

Makes things more anime-like and seems to push them into the sitting position..? Or maybe it's my imagination.

It seems impossible to make guys with this at 1.0 as well.. it drags everything into 'young anime female'

Doesn't seem to have any preference for NSFW but marking it as such just in case.

Why though?

If you don't know what this is for or why you want it - you probably don't need it.

I run into several prompts a day that use this thing and it was impossible to find..

Credits

No idea who made this or what artwork inspired it..

Trained 40+ images (fan art, renders, cosplay) for over 20k steps on SD1.5, and it's still finicky, especially for photo-realistic models. (It's like they haven't seen a woman with purple skin and horns before.) It really likes Anything-V3.

"Female draenai" should get you the best results. (Not enough source material for males.)

Now... Show Me What You Got!

Trained by 10000 CryptoPunks image data.

Use with keyword `punks_sd` like `a cropped painting of Male Hoodie Mustache Clown Eyes Green, art by punks_sd` to generate images.

Do not use it for bad things.

It is modeled after Kayano Ai, a Japanese female voice actress.

Facial close-ups have a high success rate, but the whole body tends to fall apart.

Please use LORA or other means, as the prompt alone tends to change the entire face for clothing that does not fit her image (e.g., gothic lolita clothing).

See the prompts in the sample image for usage.

First iteration of a hypernetwork trained on art by SaltyXodium, roughly 3K images, network structure [1, 1.25, 1.25, 1.25, 1.25, 1], no dropout, no normalization, LN 1e-5.

Recommended negative prompt keywords: author name, text, signature, border, outside border

Hypernetwork strength above 0.90 should be avoided. See comparison charts shown on HN 55k steps and 5K steps.

Model: AnythingV3 (training and sample images)

azuki hypernetwork,trained by anything v3

I made 2 models 1 version does more similar to it. 2 the model differs from the first one, it does not always do it well, especially for weight growth, but the skin is smooth here and very useful for a beautiful cover.

Model trained on the basis of Deliberate 2.

Attention: In the promt, write the correct model name . In the script, replace <hypernet:CristyRen4:1> on <hypernet:cristyRen_v20:1>. Also for the first version, replace <hypernet:CristyRen3:1> on <hypernet:cristyRen_v20:1>

Also, to improve the picture, use Hi res. fix, Hires steps is enough to leave for 15

For the sake of interest, I trained the hypernetwork on the same dataset as LORA. The result was extremely interesting and beautiful.

I used the same promts for clarity.

In general, the hyper network works well with values of 1, but also with values of 0.5 is also not bad.

Heya! This model works extremely well and was delicately trained to create everything in the style of the Entergalactic movie by Kid Cudi.

BASED ON MEINAV7.

IT WILL ONLY WORK ON THIS EXACT CHECKPOINT.

Other than that, have fun and be sure to use good negative prompts.

正面身体还是不错的

First time postin here.

I trained this model with hard shadows and colorful lighting in mind. It also is quite good at creating backgrounds and detail shots of objects.

Use with CakemixV10.

I TRAINED IT USING THIS MODEL, SO ONLY THIS ONE WILL WORK.

ALSO BE SURE TO USE GOOD NEGATIVE PROMPTS.

Have fun, and be sure to send me your creations.

I did not make any of the hypernetworks, all credit to original creator name(hypernetworks).
the full permissions of the model goes to the creator.

Half Korean and half Dutch (Canadian) solo Kpop star Jeon Somi.

Highly recommend "Caucasian woman". Without "Caucasian" all models I checked exaggerate her Korean features.

Also good for adding a bit of Wasian flavour to your girls using lower weights.

Works well with Deliberate. Realistic Vision if you want a bit stronger features. A mix of the two works even better for likeness. Chillout Mix works as well with "realistic" added (see angel prompt).

About 1 in 5 images should be a good likeness. The other girls are cute too, especially in Deliberate.

Hypernetwork trained on 600 unique images (1 200 after mirror) with a focus on faces (portraits). Low res images were upscaled, all images were stretched to 512x512, tagging done by "WD1.4 tagger" automatic1111 extension.

All images produced with AnythingV3 model + its VAE

This is a hypernetwork trained to produce anthropomorphic deer characters that look like Noelle Holiday from Deltarune.

0.7 on Anything is producing decent results for me.

Russian speaking twitch streamer, lets player and vlogger.

Trigger word: young woman

Examples generated on the Deliberate model.

I recommend that you first generate the desired image, and then change the face using hypernetwork in the inpaint tab:

Inpaint settings:

1. Resize mode - Just resize

2. Mask mode - Inpaint masked

3. Masked content - original

4. Inpaint area - Only masked

5. Width and Height - 512x512 or higher

6. Don't forget to switch to a random seed

7. Experiment with Denoising strength (0.3 - 0.9) and Only masked padding, pixels (10-60)

Yuumi

This is a hypernetwork trained to create artwork of Gris, an original character of Vertigris.

My AI album, more coming soon!

This is a hypernetwork trained to produce characters that would resemble the art style of Gatodelfuturo (gaturo).

The recommended hypernetwork strength is 0.65-1.0

Anything 4.5 seems to be making decent things with this.

P.S. мирзаев, иди нахуй

This is a hypernetwork that makes it possible to generate Teri, the paper bear from The Amazing World of Gumball series.

My personal best results were achieved on Anything 4.5, with hypernetwork strength around 0.75-0.95.

An experimental hypernetwork trained on 100+ different anime characters. Do we actually need to create specific networks for each and every character or could we work on improving our base models instead?

Since the list is rather large, see each version for which characters were added. Later versions include all characters of the previous versions as well.

This kind of model certainly requires a lot more training than just doing a single character and some characters might not be as well finished as others. However adding more characters does not negatively affect previously trained characters, indicating that the network have no issues representing them and that it is simply an issue of training it enough.

This is the model of ancient style illustration. The first experimental training model has many shortcomings, which are more inclusive.

I can't do too well because of my limited ability. I also hope that some big men will continue in this direction.

There are no keywords, just enable them directly in the settings,

Mainly in cooperation with Novelai,

other anime painting styles can also be integrated

这是古风插画的模型，第一次实验训练模型有很多不足，多多包涵。

能力有限没法做太好，也希望有大佬在这个方向上继续下去。

也没什么关键词之类的，直接在设置中启用即可，

主要是配合Novelai的，其他二次元画风也能融合一部分

也希望有大佬能制作下国潮插画，国潮建筑之类的

Hypernetwork for making cute and funny dollfie like anime face with realistic/semirealistic body and photographic images.

I create this hypernetwork on late january for personal fun and has been playing with it to create dollfie images and i think i will share this with you all.

Feel free to post your "cutest" image on the discussion section, i want to see it 😉.

trained on AbyssKotosMix (https://civitai.com/models/5245/kotosmix), i suggest you to use the same models. i think it also can be used with any anime based models (like anythingv3), also compatible with some character LORA.

Focus trained on dataset images of DDH-01, DDH-09, DHH-10, DHH-22, DHH-27, DDH-28 and my own dolls (mainly with cute face), 2000steps.

I will improve this hypernetwork and maybe create LORA version of it in the future.

I suggest to use cfg scale : 7 or lower.

This hypernetwork also good at img2img

Recomended prompt : 1girl, 3d, realistic, bangs, depth_of_field, eyelashes, motion_blur, solo, depth_of_field,bokeh, chromatic_aberration, (highres:1.05), (intricate detail:1.1025)

for negative just use the standard negative prompt or use EasyNegative https://civitai.com/models/7808/easynegative

example images used kotosmix at clip skip 2 as well as the Textual Inversions, bad-artist, bad-prompt-2, and bad-hands-5

I recommend setting the hypernetwork strength to 0.9

2nd image is a custom character i trained

Watercolour hypernetwork that is pretty useful for generating portraits with the soft painterly look on it. Can be used with both anime and realistic(maybe) models.

Would recommend testing to see what weights work for you. it might be a little too strong on anime models? might need to use lower weights for those while 0.8-1 on realistic models should be fine

Trained on mostly portraits so it might not be able to do backgrounds that well or default to background that is a splash of colour. For some reason likes to sprout flowers even unprompted.

Trained on the same data as the NEON ISOMETRIC model - but comes up with some vibrant different results as per the images

Coffee is nice: https://ko-fi.com/DUSKFALLcrew

--no joke, coffee is amazing, so if YOU DO like what we do, PLEASE send us some money for our bad addictions.

Hypernetwork for other model

hypernetwork of actress Eva Green. triggerword (not that important with hypernetworks tough) is 3v4gr33n, I suggest adding 'green' to your negative prompt

probably a bit overcooked, I suggest not using it at full strength

Hypernetwork for the model of the same name.

Can help generate better images.

Hypernetwork that creates gigantomastia breasts. Works ok with Uberrealistic porn merge v1.2 and Real BB 3.0. I've used clip skip 2 (for some reason) and ~0.8 multiplier for ok'ish results. Kind of needs some Lora's in conjunction. Use "gigantic breasts close up gigantomastia" as part of prompt.

It's the first thing I've trained so sorry for bad quality.

a hyper network trained by よし男's artwork.

(reference: https://www.pixiv.net/users/3584828)

only for study and self use

please do not publish or use for business.

请勿发表或商用

Author: Tongfan Wei (weitf@bu.edu)